On Fri, 19 Jan 2007, Philip Hazel wrote:
> I agree that an approximation could well be useful, but I am sure that
> people would expect a variable called connection_count to be the real
> connection count. SMTP sessions can easily last for minutes, so the
> value could be very wrong by the time it is used. That's why I want to
> make sure the name indicates that there is something anomalous about
> this value.
I have committed a patch that implements $smtp_count_at_connection_start
so it should be in tonight's snapshot. Documented in doc/NewStuff.
Philip
--
Philip Hazel, University of Cambridge Computing Service.
